AgenticAKM : Enroute to Agentic Architecture Knowledge Management

2026-02-04

AI Summary

This work addresses the persistent challenges in software Architecture Knowledge Management (AKM), which stem from the wide distribution of knowledge and its strong contextual dependencies. To tackle these issues, the paper presents the first large language model–based multi-agent system that decomposes AKM into four coordinated subtasks—extraction, retrieval, generation, and validation—each handled by a specialized agent. This collaborative approach effectively mitigates the problems of fragmented architectural knowledge and context limitations. Evaluated through user studies across 29 real-world code repositories, the system automatically produces high-quality Architecture Decision Records (ADRs), significantly enhancing both the automation and practical utility of AKM. The results demonstrate the novelty and effectiveness of the proposed multi-agent framework.

Technology Category

• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Multiagent Systems: Agent/AI Theories and Architectures
• Data Mining & Knowledge Management: Knowledge Acquisition from the Web

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Architecture Knowledge Management (AKM) is crucial for maintaining current and comprehensive software Architecture Knowledge (AK) in a software project. However AKM is often a laborious process and is not adopted by developers and architects. While LLMs present an opportunity for automation, a naive, single-prompt approach is often ineffective, constrained by context limits and an inability to grasp the distributed nature of architectural knowledge. To address these limitations, we propose an Agentic approach for AKM, AgenticAKM, where the complex problem of architecture recovery and documentation is decomposed into manageable sub-tasks. Specialized agents for architecture Extraction, Retrieval, Generation, and Validation collaborate in a structured workflow to generate AK. To validate we made an initial instantiation of our approach to generate Architecture Decision Records (ADRs) from code repositories. We validated our approach through a user study with 29 repositories. The results demonstrate that our agentic approach generates better ADRs, and is a promising and practical approach for automating AKM.

Problem

• Architecture Knowledge Management
• Software Architecture
• Large Language Models
• Architecture Decision Records
• Distributed Knowledge

Innovation

• Agentic Architecture
• Architecture Knowledge Management
• Large Language Models
• Architecture Decision Records
• Multi-agent Collaboration